Positive STEPS: Enhancing Medication Adherence and Achieving Viral Load Suppression in Youth Living With HIV in the United States—Results From an Efficacious Stepped Care, Randomized Controlled Trial

Abstract

Background:

In the United States and worldwide, there is a significant number of young people acquiring and living with HIV. Antiretroviral therapy (ART) has led to significant reductions in HIV-related illnesses and deaths, allowing young people living with HIV to manage their condition as a chronic disease. Ensuring high levels of ART adherence is vital for treatment success. Despite this, to the best of our knowledge, there are no efficacious behavioral interventions for improving ART adherence and viral suppression among youth in the United States.

Methods:

We conducted a multicity, randomized, controlled trial—in Boston, MA/Providence, RI and Chicago, IL—to examine the efficacy of an stepped care, behavioral and technology–based intervention, “Positive STrategies to Enhance Problem-Solving Skills (Positive STEPS),” compared with a standard of care (SOC) control, for improving ART adherence and viral suppression among youth living with HIV ages 16–29 years. Positive STEPS included the following: step 1: TXTXT, an evidence-based, daily 2-way personalized text message reminder to take ART medications; step 2: only participants with <90% adherence anytime between weeks 5 and 12 postrandomization to the Positive STEPS arm would then receive five 50-minute sessions of manualized individual adherence counseling. If participants’ adherence remained at ≥90% , then they did not progress to step 2. ART adherence was measured via Wisepill, an electronic medication monitoring device, and self-report. Participants were followed for 12 months and completed biospecimen collection (HIV plasma RNA viral load testing) and a quantitative assessment battery at baseline and at their 4-, 8-, and 12-month follow-up visits.

Results:

Between March 2018 and March 2023, 123 participants were randomized (Positive STEPS = 63; SOC control = 60). Intention-to-treat analyses showed a significant positive main effect for the Positive STEPS arm, which increased the mean log ART adherence by 18.7% relative to the SOC control at the 4-month visit (coefficient = 0.187, P = 0.021). For the time effect, a significant overall increase in ART adherence across the subsequent follow-up visits was observed, with increased mean log ART adherence by 27.8% (P < 0.01) at 8 months and 30.1% at 12 months (P < 0.01), relative to the SOC control. With respect to our viral suppression outcome, the analysis revealed a significant negative main effect for the Positive STEPS arm at 4-months (odds ratio = 0.264, P = 0.023), indicating that the odds of having an unsuppressed virus were 74% lower in the Positive STEPS arm compared with the SOC control; the interaction term revealed that this effect was maintained through the 12-month follow-up time period.

Conclusion:

These findings on the efficacy of Positive STEPS to enhance ART adherence and viral suppression among youth living with HIV represent the first behavioral intervention for youth to show significant and sustained effects on both behavioral (Wisepill and self-report) and biomedical (HIV viral load) outcomes related to ART adherence. The intervention not only demonstrated remarkable efficacy when compared with the SOC control but also showed maintenance of gains over a 12-month period. Effectiveness and implementation science approaches to testing positive STEPS in real-world settings is recommended.

INTRODUCTION

In the United States and worldwide, there is a significant number of young people acquiring and living with HIV. According to the Centers for Disease Control and Prevention, in 2022, the largest number of new HIV infections (12,700) occurred among people aged 25–34 years.¹ Moreover, people aged 13–24 years accounted for 20% (6400) of the estimated 31,800 new HIV infections in 2022. Fortunately, advances in medical treatment, particularly antiretroviral therapy (ART), have led to significant reductions in HIV-related illnesses and deaths, allowing young people living with HIV to manage their condition as a chronic disease. Nonetheless, ensuring high levels of treatment adherence is vital for treatment success, and promoting adherence remains an essential aspect of contemporary HIV care. Regardless of this, for every 100 youth aged 13–24 years diagnosed with HIV in 2021, only 55% were retained in HIV care, and 35% had unsuppressed viremia.¹

Adherence to ART is crucial for achieving optimal health outcomes in individuals living with HIV, particularly among youth who face unique challenges in managing their medication. Poor adherence to ART can lead to the development of drug resistance, disease progression, and increased risk of sexual transmission, highlighting the critical need for effective interventions to improve medication adherence in this vulnerable population. In light of these considerations, the most promising approaches to enhance treatment adherence among youth living with HIV may encompass a multi-faceted approach, incorporating elements such as patient education, self-monitoring, and medication reminders. The rationale for employing multilevel components aligns with interventions for adult adherence, as multiple strategies have shown the greatest efficacy in bolstering adherence.^2,3 However, to our knowledge, there are no published, full-scale, randomized, controlled, efficacy trials (RCT) of successful interventions designed to improve ART adherence among youth living with HIV in the United States, and none meeting the Center for Disease Control and Prevention’s best evidence threshold for strength of evidence.^4,5 There is one full-scale RCT of an intervention called “Project Yes!” conducted among youth living with HIV and aged 15–24 years in Ndola, Zambia, but there were no significant intervention effects on self-reported ART adherence, and intervention effects achieved on viral suppression were significant at the 6 month assessment but were not sustained at any follow-up visits over the 18 months of observation.⁶

Among the limited pilot RCTs of interventions conducted to enhance ART adherence among youth living with HIV, the two that are most promising were published by the authors of the this article. The first is “TXTXT,” a personalized text messaging intervention to enhance ART adherence among youth ages 16–29 years, which found that a significantly greater proportion of those in the TXTXT arm reported ≥90% medication adherence compared with those in the control arm at 3- and 6-month postrandomization.⁷ The second is called “positive Strategies To Enhance Problem-solving Skills (positive STEPS),” combining 5 individual adherence counseling sessions with personalized daily text message reminders for youth ages 13–24 years living with HIV.⁸ The pilot RCT of positive STEPS found that at the 4-month assessment visit, the change in ART adherence among the intervention arm was significantly higher compared with the standard of care (SOC) control.⁸ As these were both pilot RCTs, neither were powered to find significance on viral suppression. These 2 pilot RCTs were the major preliminary studies for testing the efficacy of positive STEPS in the full-scale RCT presented in this article.

Of the other pilot trials published in the literature, 2 necessitated substantial study resources and staff time to carry out home visits and monitor directly observed therapy, raising concerns about sustainability in real-world settings.^9,10 Additionally, other pilot studies focusing on ART adherence interventions for youth and young adults living with HIV addressed singular barriers to adherence, such as pill swallowing¹¹ or the implementation of a medication reminder system.¹² Furthermore, 6 additional prior pilot trials of interventions aimed at enhancing ART adherence among youth and young adults have incorporated education and skill building and yielded modest effects.^5,13–18 Because, to our knowledge, there are no published full-scale RCTs with established efficacy among youth, it is imperative to develop interventions that are not only acceptable and feasible to implement but also responsive to the distinct challenges in ART adherence and the contextual realities faced by youth living with HIV.

In this study, an innovative stepped care model—positive STEPS¹⁹—was employed to enhance ART adherence among youth living with HIV. Our model prioritizes the delivery of the least resource-intensive intervention initially, with only those who do not show improvement receiving the more resource-intensive intervention component.¹⁹ This streamlined approach is purposefully designed to allocate resources judiciously while ensuring that participants receive effective interventions. The current article outlines the methodology for evaluating efficacy and presents findings of a full-scale RCT on the efficacy of the positive STEPS intervention compared with a SOC control in achieving viral suppression and enhancing monitored (ie, Wisepill) adherence among youth living with HIV aged 16–29 years.

METHODS

Study Design and Participants

Study Design

We conducted a 2-arm RCT from March 2018 to March 2023 to examine the efficacy of a behavioral and technology-based intervention, “positive STEPS” (intervention arm), compared with a SOC control arm, for improving ART adherence and viral suppression among youth living with HIV ages 16–29 years. The positive STEPS intervention included (1) step 1: daily 2-way personalized text message reminders to take ART medications and (2) step 2: for participants with <90% adherence anytime between weeks 5 and 12 postrandomization to the positive STEPS arm would then receive five 50-minute sessions of manualized individual cognitive behavioral adherence counseling using the adolescent adapted version of LifeSteps, called “positive STEPS.”^19–21 We monitored ART adherence among participants in both arms using Wisepill technology, which makes use of cellular and internet technologies, allowing researchers to access participant data in real time, on a daily basis using the manufacturer’s website. During the baseline visit, participants are given a Wisepill device and shown how to use it. The total duration of participation was 12 months with 4 major assessment visits at baseline, and at 4- (acute outcome), 8-, and 12-month postrandomization (Fig. 1). This trial is registered with ClinicalTrials.gov: NCT03092531.

FIGURE 1.

Study consort diagram for positive STEPS randomized controlled efficacy trial (N = 123 randomized). Note: *Those who were withdrawn or LTFU were not significantly different in terms of baseline sociodemographic and adherence characteristics. **25 participants (pts.) completed all 5 adherence counseling sessions; 3 pts. completed 4 sessions each; 1 pt. completed 3 sessions; 1 pt. completed 2 sessions. LTFU, lost to follow-up.

Participants

In total, 123 participants were randomized (SOC control = 60 and positive STEPS = 63); study CONSORT diagram illustrates participant retention of both assessment visits and positive STEPS counseling sessions (Fig. 1). Inclusion criteria were youth living with HIV ages 16–29 years who had been prescribed ART for at least 3 months. Additional inclusion criteria were self-reported difficulties with ART adherence (missed ≥4 doses in the past month or ≥1 dose in the past week), ownership of/daily access to a mobile phone, and having lived in Providence/Boston or Chicago areas for ≥3 months. Participants were excluded if they reported a past year intervention or currently participating in an intensive intervention for medication adherence, unable to consent due to illness/cognitive limitations/intoxication, and planning to move outside of the study areas within 1 year. A total of 683 participants were screened for the study; 227 met inclusion criteria, 143 were enrolled, and 123 were randomized (Fig. 1).

Eligible participants provided written informed consent before being enrolled. Institutional Review Board approval was granted at the Brown University (Providence, RI), The Fenway Institute (Boston, MA), and Lurie Children’s Hospital (Chicago, IL).

Study Setting and Recruitment

The study was conducted at 2 sites (3 US cities): site 1: Providence, RI/Boston, MA, and site (2) Chicago, IL. Participants were recruited from HIV clinics, university medical centers, AIDS and youth-focused service organizations, various youth events (eg, youth pride), and through social media (Facebook, Instagram, GRINDR, etc).

Randomization and Masking

All participants completed a Wisepill “run-in” period 2 weeks following their baseline/enrollment visit and then a randomization visit at the end of this period. Participants were block randomized with equal allocation to positive STEPS and the SOC control.

Alternating block sizes of 4–8 were used to ensure balanced assignment and so study staff could not guess assignment to either arm. Randomization was performed by computer-generated allocation sequences. Study staff were completely masked to the randomization until the completion of the visit.

Procedures

Description of the Positive STEPS Intervention and SOC Control

All participants, including those assigned to positive STEPS and the SOC control, were provided information about youth and young adult mental health services, drug/alcohol treatment, and/or other adolescent HIV services in their area.

Positive STEPS

The “positive STEPS” intervention has 2 components: (1) text messaging and (2) adherence counseling. The overall intervention package is anchored in the social cognitive and contextual realities of young individuals living with HIV,²⁰ encompassing a comprehensive understanding of their living circumstances, familial support, vocational engagements, and educational pursuits. This approach is underpinned by the tenets of social cognitive theory,²² which delineates a core set of mechanisms that shape health behavior, with a primary focus on self-regulation—comprising the precision and constancy of self-observation and self-monitoring—and self-reflection, including the cultivation of self-efficacy, denoting the belief in one’s capacity to exert control over the targeted health behavior. The reason for having 2 distinct components was to allow for an efficient stepped care (“adaptive”) model²³ in which the least resource-intensive intervention (text messaging) is delivered first, and only those individuals who do not improve, then receive the higher-intensity, more resource demanding intervention (adherence counseling). For step 1 of the stepped care model, participants receive lower-intensity, daily, personalized, 2-way texts messages for 12 months; these messages serve as social cognitive cues or reminders to take their medications according to their medication schedule. A structured interview was used to personalize the messages to the participant’s needs, including the timing of dosage(s) and content of the message (eg, to focus on reasons why taking HIV medications as prescribed is important to each individual). If participants demonstrated ≥90% adherence—monitored via Wisepill—they remained on step 1.

Participants who did not demonstrate adequate adherence to their HIV medications (<90% adherence) anytime up to the end of month 3 (weeks 5–12) progressed to step 2 of the stepped care package–adherence counseling. The counseling involved 5 in-person cognitive behavioral sessions, each lasting approximately 50 minutes, following a manualized approach.^8,20–22 The counseling approach was based on LifeSteps²⁴ but adapted so that sessions were adolescent specific, included digital video vignettes, and was delivered by master’s-level counselors. During each counseling session, the participant and counselor went through a series of potential adherence problems and challenges, and when relevant, generated alternative solutions, made decisions about the alternatives, determined the optimal solution, and developed an action plan (or updated it) with skills training to implement that solution. Each session began with a discussion of participant’s adherence over the prior week (since their last visit) using a motivational interviewing style. When appropriate, the counselor engaged in a discussion with the participant about how drugs and/or alcohol may interfere with adherence and problem solving situations involving drug and/or alcohol use, establishing a plan to manage these challenges, and providing referrals to substance abuse treatment. During the sessions, the counselor and participant discussed how substances are used (ie, frequency, context, setting, alone/with friends/with sexual partners/others, other’s use around them) and determined if substance use has been a barrier to optimal ART adherence for the participant.

Standard of Care Control

The SOC control received standard health services offered to all patients in the context of their HIV care (eg, mental health services, case management) and a brief adherence educational session, which included a 20-minute animated tutorial that explained the importance of adherence to antiretroviral medication effectiveness, toxicity expectations, and medication misperceptions. During the educational session, participants had the opportunity to ask questions and received answers to reinforce their current ART regimen. This video is specifically designed to be viewed by individuals who have no scientific background and is appropriate for youth and young adults. The video, which is titled, “HIV Drug Resistance and the Importance of Adherence,” was created by BioCreations, a digital media company, in collaboration with the Johns Hopkins Point of Care Information Technology Center. The educational session was implemented during the randomization visit among those assigned to the SOC control.

Fidelity Monitoring

The integrity of the specific protocols were assured empirically. All positive STEPS counseling sessions were audio-taped. We developed a rating checklist for counselor adherence that includes whether the specific components of each session was, in fact, delivered. Adherence rating/fidelity monitoring was completed by the clinical supervisor at both study sites. At least 20% of the sessions, and at least 1 session per week, was reviewed before clinical supervision, and the checklist was used to give feedback to the counselors.

Outcomes

Study Assessments

Participants completed quantitative assessments administered through an iPad using an audio computer-assisted self-interview system at baseline and at 4-, 8-, and 12-month follow-up visits.

Primary Outcomes

Our primary outcomes for this study were ART adherence and viral suppression. Adherence measures included data from Wisepill device and self-reported adherence.

Wisepill Adherence

Wisepill is an innovative electronic medication monitoring device that is used to characterize patterns of ART adherence over the study period for both study arms. It is designed to be discrete and easy to transport.²⁵ Whenever participants opened the device, it triggered an “intake” signal to a central server, indicating a dose was likely taken. A “heartbeat” signal occurred when the device was working, but it was not opened. When the device was malfunctioning, it sent a “none” signal. We received daily adherence information from each participant’s Wisepill device. A missed dose was coded with 0, and a dose taken was coded as 1. Percent adherence was calculated by dividing the number of times the device was opened by the expected number of openings per ART prescription. In analyses, periods when devices malfunctioned were considered missing data.

Self-reported Adherence

Each participant was asked to recall days they took their ART medication doses in the preceding 30-day and 4-month periods.^8,19 Specifically, participants were asked, “In the last month, what percent of the days did you take your HIV medication doses?” and “In the last 4 months, what percent of the days did you take your HIV medication doses?”.

Viral Suppression

At all major study assessment visits (baseline and 4, 8, and 12 months), a single tube of blood was collected for testing HIV plasma RNA.²⁶ Participant’s viral load was dichotomized as virally suppressed (<200 copies/mL) or virally unsuppressed (≥200 copies/mL) for each assessment visit.

Sociodemographic Characteristics, and Psychosocial and Cognitive Behavioral Factors

The validity, reliability, and scoring of the following measures have been previously described in the published protocol.¹⁹ At baseline, we collected sociodemographic questions from each participant, including age, sexual and gender identity, sex assigned at birth, race/ethnicity, study site, educational attainment, and student/employment status. We also collected data on the number of years living with HIV and prescribed ART. In addition, we assessed clinically significant depressive symptoms, through the 20-item Center for Epidemiologic Studies Depression Scale¹⁹ and clinically significant symptoms of anxiety, assessed using the Beck Anxiety Inventory (BAI).¹⁹ Problem alcohol and drug use were also assessed with the Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST).¹⁹ The ASSIST consists of 9 items, covering 10 substances (used in the past 4 months), including tobacco, alcohol, cannabis, cocaine, stimulants, inhalants, sedatives, hallucinogens, opioids, and other drugs. The ASSIST assesses frequency of use and associated problems for each substance with good to excellent reliability and validity. We measured participants’ social support, which was assessed using the ACTG outcomes committee assessment and measures social support for adherence.¹⁹ The HIV Medication Taking Self-Efficacy Scale was used to measure ART adherence self-efficacy or one’s confidence in their ability to take HIV medication in various situations.¹⁹ Motivation and outcome expectancies of ART adherence was assessed as 3 separate dimensions: attitudes, norms, and behavioral intentions to adhere to HIV medication.¹⁹ Finally, self-regulation skills, which included self-monitoring, goal-setting, and enlistment of self-incentives/plans, were also assessed.¹⁹

Statistical Analysis

All data were analyzed using R version 2023.12.0 + 369. To reduce the threat of bias, the intent-to-treat principle was used for all analyses, where participants were analyzed according to randomization arm.

Power Analysis

The primary power analysis was calculated in R using the “pwr” package and was based on the acute ART adherence outcome (differences between the Positive STEPS and SOC control arms) from our pilot RCT.⁸ Careful consideration was given to the effect size to use in the power analysis. The pilot RCT demonstrated a significant difference in means in ART adherence between the positive STEPS and the SOC control from baseline to follow-up (P = 0.028; effect size: 0.612). For these estimates, we used a 1-sided test given our hypothesis of positive intervention effects and alpha = 0.05.

Group sizes of 60 randomized per arm (N = 120 total) will give >80% power to detect at least a 10% or greater difference in adherence at 4 months. Moreover, based on our preliminary studies in youth living with HIV—for our binary viral suppression outcome—we will have >80% power to detect a 15% difference. These calculations assume that the proportion of subjects with viral suppression in the SOC control is 0.5, the type I error rate is 5%, and the correlation between observations on the same subject is 0.5.

Baseline Sociodemographic, HIV, and Psychosocial Related Characteristics

Means and SD were calculated for continuous baseline characteristics, and categorical data were summarized using proportions. Differences in baseline sociodemographic and psychosocial factors, as well as HIV treatment history questions were compared between the positive STEPS and SOC control arms using t test and χ² tests as appropriate (Table 1).

TABLE 1.

Baseline Sociodemographic Characteristics and Psychosocial Problems Among Youth Living With HIV by Randomization Arm (N = 123)

	Total Sample (N = 123)		Positive STEPS (n = 63)		SOC Control (n = 60)
	Mean	SD	Mean	SD	Mean	SD	p†
Age in years (range 19–29)	26	2.9	25.1	2.9	25.8	2.9	0.129
Years living with HIV‡	6.0	6.1	6.1	5.9	6.0	6.4	0.671
Years prescribed ART§	6.0	7.4	6.0	7.3	6.0	7.5	0.839
Social support‖ (range 1–5)	4.0	1.2	4.2	1.1	3.8	1.3	0.204
Adherence self-efficacy¶ (range 1–10)	7.6	2.0	7.6	2.0	7.6	2.0	0.951
Outcome expectancy# (range 1–10)	8.7	1.9	8.7	1.7	8.7	2.0	0.768
Motivation** (range 0–2)	1.6	0.4	1.6	0.5	1.7	0.3	0.060
Self-regulation (range 1–5)	3.8	0.4	3.8	0.4	3.9	0.3	0.245
	N %		N %		N %		P †
Sex assigned at birth
Female	15	12.2	8	12.7	7	11.7
Male	108	87.8	55	87.3	53	88.3	1.000
Sexual identity
Straight heterosexual	14	11.4	7	11.1	7	11.7
Gay/homosexual	71	57.7	35	55.6	36	60.0
Bisexual/other	38	30.9	21	33.3	17	28.3	0.834
Enrollment site
Chicago	76	61.8	38	60.3	38	63.3
Providence/Boston	47	38.2	25	39.7	22	36.7	0.874
Race/ethnicity
Black/African American	65	52.8	33	52.4	32	53.3
White	25	20.3	10	15.9	15	25
Hispanic/Latinx (White)	18	14.6	9	14.3	9	15.0
Other	15	12.2	11	17.5	4	6.7	0.244
Educational attainment
High school diploma/GED	45	36.6	22	34.9	23	38.3	0.382
Some college	34	27.6	20	31.7	14	23.3
College	22	17.9	10	15.9	12	20.0
Some graduate/graduate degree	8	6.5	2	3.2	6	10.0
Other	14	11.4	9	14.3	5	8.3
Employment
Unemployed	49	39.8	30	47.6	19	31.7	0.105
Employed full/part time	74	60.2	33	52.4	41	68.3
Student status
Enrolled in school	29	23.6	13	20.6	16	26.7	0.565
Not enrolled in school	94	76.4	50	79.4	44	73.3
Clinically significant depressive symptoms (CES-D)††
No (score < 16)	41	34.2	20	32.8	21	35.6	0.895
Moderate (score = 16–24)	36	30.0	17	27.9	19	32.2
Severe (score ≥ 25)	43	35.8	24	39.3	19	32.2
Anxiety symptoms (beck anxiety inventory)‡‡
Minimal anxiety (score 0–7)	8	6.7	5	8.1	3	5.2	0.250
Mild anxiety (score 8–15)	40	25.8	18	29.0	22	37.9
Moderate anxiety (score 16–24)	47	33.3	22	35.5	25	43.1
Severe anxiety (score 25–63)	25	20.8	17	27.4	8	13.8
Tobacco use—past 4 mo
No	57	46.3	32	50.8	25	41.7	0.404
Yes	66	53.7	31	49.2	35	58.3
Heavy alcohol use—past 4 mo (ASSIST) (range 0–39)
Lower risk (scale < 11)	89	72.4	47	74.6	42	70.0	0.712
Higher risk (scale ≥ 11)	34	27.6	16	25.4	18	30.0
Polydrug use*—past 4-mo
No	101	82.1	50	79.4	51	85.0	0.562
Yes	22	17.9	13	20.6	9	15.0
Cannabis use—past 4 mo
No	32	26.1	16	25.4	16	26.7	1.000
Yes	91	73.9	47	74.6	44	73.3
Cocaine use—past 4 mo
No	98	79.7	47	74.6	51	85.0	0.227
Yes	25	20.3	16	25.4	9	15.0
Amphetamine use—past 4 mo
No	99	80.5	50	79.4	49	81.7	0.925
Yes	24	19.5	13	20.6	11	18.3
Inhalants/poppers use—past 4 mo
No	107	87.0	54	85.7	53	88.3	0.870
Yes	16	13.0	9	14.3	7	11.7
Opioids use—past 4 mo
No	119	96.7	61	96.8	58	96.7	1.000
Yes	4	3.3	2	3.2	2	3.3

^*Polydrug use = .3 drugs used (cannabis, cocaine, amphetamines, inhalants/poppers, hallucinogens, or opioids).

^†Statistical tests are assessing differences at baseline between the Positive STEPS arm vs the SOC control; P values were obtained using the χ² tests for categorical variables and the Wilcoxon rank-sum test for continuous variables.

^‡Participant (1) had missing information on the number of years living with HIV.

^§Participants (2) had missing information on the number of years using with ART.

^‖Participants (2) had missing information on social support.

^¶Participant (1) had missing information on self-efficacy belief.

^#Participant (1) had missing information on outcome expectancy.

^**Participant (1) had missing information on motivation.

^††Participants (3) had missing information on CED-D scale.

^‡‡Participants (3) had missing information on BAI scale.

CES-D, Center for Epidemiologic Studies Depression Scale, GED, General Educational Development.

Missing Data

We used multiple imputation procedures to account for any missing ART adherence and viral suppression data. Multiple imputation models included key sociodemographic and clinical variables that were associated with missingness for each outcome. The percent adherence and viral suppression analyses were performed separately on each imputed data set. To account for missing Wisepill data during periods of devise malfunctioning (ie, not connecting to the remote server), we imputed these data using the self-reported ART adherence assessment collected at each study visit (4, 8, and 12 months). To fully allow for intent-to-treat principles, we conducted Generates Multivariate Imputations by Chained Equations (with 20 imputations) to provide conservative estimates for missing data. This was performed using the Multivariate Imputation by Chained Equations package for R.²⁷ Analyses that did not impute missing values revealed the same pattern of results. At the 12-monh follow-up visit, retention was 84%, equating to 16% missingness.

Primary Analysis

The primary analysis compared differences in percent adherence and viral suppression status at 4-, 8-, and 12-month visits between the randomized arms. To ensure a more accurate comparison, we excluded baseline data. This exclusion was necessary because participants began using the Wisepill device at enrollment, precluding baseline ART adherence measurements. There were no significant differences between randomization arms regarding baseline adherence or viral suppression. For all analyses, statistical significance was evaluated at an alpha level of 0.05.

Primary Outcome: ART Adherence

ART adherence was defined as the percentage of doses taken as prescribed, monitored through Wisepill. Adherence was analyzed over 3- to 4-month intervals within a year using linear mixed-effects models. The logarithm of the measured ART adherence percentage was used as a continuous outcome. The model included fixed effects for treatment arm and time. The coefficient for the treatment arm represents the difference in adherence between the 2 arms, whereas the time coefficient indicates changes in adherence across successive time points. An interaction term between treatment arm and time was included to examine how the positive STEPS arm’s effect on ART adherence varied over time. Random intercepts for participants were included to account for within-subject correlation over time.

Primary Outcome: Viral Suppression

Viral load suppression was defined as a binary outcome. We analyzed using generalized linear mixed-effects models with a logit link function. Participant-level random intercepts were included in the viral suppression model to account for individual variability. This model featured terms for treatment arm, time, and time-by-treatment interaction.

Role of the Funding Source

The funder of the study—the National Institutes of Health—had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript.

RESULTS

Participant Characteristics

The overall study population included 123 youth living with HIV who were randomized to 1 of the 2 study arms; 61.8% were recruited in Chicago, and 38.2% in Providence/Boston. The mean participant age was 26 years (SD = 2.9). On average, participants had lived with HIV for 6 years (SD = 6.1) and taken ART for 6.0 years (SD = 7.4).

Clinically significant depression symptoms based on the Center for Epidemiologic Studies Depression Scale cutoff score were prevalent (65.8%). Anxiety symptoms based on the BAI cutoff score were “minimal” (6.7%), “mild” (25.8%), “moderate” (33.3%), and “severe” (20.8%). Approximately 17.9% reported 3 or more drugs used in the 4 months before baseline assessment (Table 1).

Participant demographic and clinical characteristics are shown in Table 1. Study arms did not differ on baseline sociodemographic or clinical characteristics assessed. Among the overall study population, most identified as cisgender male (87.8%) and gay regarding sexual orientation (57.7%). The sample was racially/ethnically diverse, with 52.8% of participants self-identified as Black/African American, 20.3% were White, 14.6% were Hispanic/Latino, and 12.2% were other races/ethnicities, including American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, and some identifying as “other racial/ethnic groups.” Over one-third had a high school-level education (36.6%), 27.6% reported some college education, 17.9% reported graduate-level education, 6.5% reported some graduate school or graduate degree, 11.4% reported other education levels, which include sixth–eigth grade, GED program, and technical or vocational school. More than half (60.2%) were employed full/part time, whereas 39.8% were unemployed/disabled (Table 1).

Adherence to ART (Wisepill)

A significant positive main effect was found for the positive STEPS arm, which increased ART adherence by 18.7% relative to the SOC control at the 4-month visit (coefficient = 0.187, P = 0.021). For the time effect, significant increase in ART adherence were observed at both the 8- and 12-month visits compared with the 4-month visit (coefficients = 0.278 and 0.301, respectively; P < 0.01 for both). These results suggest a significant overall increase in ART adherence across all subsequent follow-ups, with increases of 27.8% at 8 months and 30.1% at 12 months (Table 2).

TABLE 2.

Linear Mixed Models^* Assessing Adherence^† to ART Among Youth Living With HIV by Randomization Arm (N = 123)

	Coefficient (SE)	P
Time
4 mo	Reference
8 mo	0.278 (0.063)	<0.001
12 mo	0.301 (0.063)	<0.001
Group
SOC control	Reference
Positive STEPS	0.187 (0.081)	0.021
Time and group interaction
1	Reference
2	−0.098 (0.088)	0.267
3	−0.117 (0.088)	0.184
Random effects
Participant variance	0.023 (0.151)

^*All models included time, measured in 4-month intervals, randomization arm, and the interaction between time and randomization arm. Random effects for each patient are accounted for.

^†Adherence was defined using Wisepill data as the percentage of the actual number of times participants opened their ART medication device to the expected number of openings over each 4-month interval. This percentage was supplemented with self-reported data to accommodate the few instances (no more than 2 weeks in duration per participant) when the device malfunctioned (was offline). The logarithm of the measured ART adherence percentage was used as a continuous outcome.

SE. standard error.

A nonsignificant interaction between time and treatment arm was observed at both the 8-month visit (coefficient=−0.098, P = 0.267) and the 12-month visit (coefficient=−0.117, P = 0.184). The nonsignificant negative coefficients suggest that although ART adherence improved significantly over time for the positive STEPS arm, the rate of improvement due to positive STEPS did not significantly differ from the SOC across follow-up periods. In other words, positive STEPS had a significant initial effect on adherence (which was sustained overtime), but subsequently both arms showed similar improvements over time (although it remained higher than in the SOC control, maintaining a consistently positive impact). Ultimately, the data show that adherence in the positive STEPS arm was significantly higher, demonstrating an overall improved ART adherence compared with the SOC control (Table 2).

Viral Suppression

Viral suppression was defined as either unsuppressed (>200 copies/mL) or suppressed according to assay thresholds. The analysis revealed a significant main effect for the positive STEPS arm at the 4-month visit (odds ratio [OR] = 0.265, P = 0.023), indicating that the odds of having unsuppressed virus were significantly lower in the positive STEPS arm compared with the SOC control. This suggests a reduced likelihood of unsuppressed virus among those randomized to the positive STEPS arm. The effect of time on viral suppression status was not significant, with an OR of 0.543 (P = 0.276) at the 8-month visit and an OR of 0.999 (P = 0.998) at the 12-month visits. These results indicate that those randomized to the SOC control did not show significant improvement in viral suppression over the 12 months of follow-up, remaining essentially unchanged from initial levels (Table 3).

TABLE 3.

Generalized Linear Mixed Models^* of the Odds of Having Unsuppressed Virus^† by Randomization Arm Among Youth Living With HIV (N = 123)

	OR (SE)	P
Time
4 mo	Reference
8 mo	0.543 (0.562)	0.276
12 mo	0.999 (0.604)	0.998
Group
SOC control	Reference
Positive STEPS	0.265 (0.584)	0.023
Time and group interaction
1	Reference
2	1.843 (0.714)	0.392
3	1.512 (0.758)	0.585
Random effects
Participant variance	1.317 (1.148)
Interaction	Positive STEPS	SOC Control
4 mo	0.265	1.000 (reference)
8 mo	0.264	0.543
12 mo	0.384	0.999

^*All models included time (at 4-, 8-, and 12-month visits), randomization arm, and the interaction between time and randomization arm. Reference is “detectable” viral loads.

^†Viral suppression was defined based on viral load measurements: Viral loads <200 copies per mililiter were categorized as “undetectable,” indicating effective viral suppression. Viral loads more than 200 copies per mililiter was categorized as “detectable,” indicating a lack of viral suppression. In the model, “detectable” is used as the reference category.

SE, standard error.

Incorporating the interaction term revealed that at the 4-month visit, the odds of having an unsuppressed virus were much lower in the positive STEPS arm (OR = 0.265) compared with the SOC control (OR = 1.0). At the 8-month follow-up visit, the odds of unsuppressed virus remained lower in the positive STEPS arm (OR = 0.264) versus the SOC control (OR = 0.543). Similarly, at the 12-month visit, the odds of unsuppressed virus for the positive STEPS arm (OR = 0.384) persisted, which was substantially lower compared with the SOC control (OR = 0.999). The nonsignificance of the interaction term implies that the treatment effect does not vary significantly over time, meaning the relative difference in viral suppression odds between positive STEPS and SOC remained fairly constant across follow-up visits (ie, over the 12-month follow-up period) (Table 3). Overall, these findings demonstrate that the positive STEPS intervention consistently reduced and maintained the odds of having an unsuppressed virus compared with the SOC control across all follow-up visits.

DISCUSSION

The findings from this study provide valuable insights into the impact of the positive STEPS intervention on ART adherence and viral suppression outcomes among youth living with HIV over a 12-month period. The results demonstrate a significant and sustained positive main effect of positive STEPS on increasing ART adherence levels compared with the control arm. These results suggest that the positive STEPS intervention has the potential to yield long-lasting benefits in enhancing ART adherence among youth living with HIV. In addition to improvements in ART adherence, the findings also highlight a significant negative main effect of the positive STEPS intervention on the unsuppressed viral outcome over a 12-month period. The odds of participants in the positive STEPS arm having an unsuppressed virus were significantly lower compared with those in the control at 4 months. Importantly, this effect was maintained throughout the 12-month follow-up period, indicating the sustained impact of the positive STEPS intervention on viral suppression among youth living with HIV.

Numerous studies have documented the complexities associated with medication adherence among youth living with HIV.^2,3,5 Although several interventions have been developed to address ART adherence, there remains a paucity of evidence regarding the efficacy of interventions specifically with regard to viral suppression and specifically tailored to the needs of youth living with HIV.^2,3,5 Given the unique developmental, psychosocial, and behavioral characteristics of this population, interventions must be carefully designed and rigorously evaluated to determine their impact on ART adherence. The positive STEPS intervention takes advantage of the widespread use of technology among youth by incorporating personalized text message reminders as social cues to facilitate ART adherence, drawing on previous evidence of the acceptability and feasibility of this approach.^7,8,19–21 If personalized text messaging were insufficient in addressing their adherence barriers, participants would then transition to a more intensive counseling intervention. Modeled after the efficacious Life-Steps program by Safren et al.²⁴ for ART adherence among adults living with HIV, positive STEPS is rooted in the principles of cognitive behavioral therapy, motivational interviewing, and problem solving and encompasses 11 steps addressed over 5 in-person counseling sessions conducted by a master-level counselor.²⁸ This study shows that this can be adapted to the unique needs of youth and that the stepped care approach is a viable way to get higher intensity/higher cost (eg, counseling) to those who need it.

Despite the manifold health advantages attributed to ART, sustaining consistent long-term adherence to treatment poses a formidable obstacle for youth coping with HIV. Alarmingly, the prevailing rate of ART adherence among youth frequently falls short of the optimal standard. A comprehensive review encompassing 22 published studies⁵ unveiled that adherence to ART among youth ranged from 28% to 69%, notably below the critical threshold of 85%–95% essential for maximizing therapeutic efficacy. Furthermore, a large-scale investigation monitoring youth living with HIV in the United States disclosed that merely 41% of those receiving ART reported adherence levels surpassing 95%.²⁹ Influential factors affecting ART adherence, such as adhering to a consistent daily dosing regimen and averting missed doses stemming from disruptions in routine, persist as substantial challenges, particularly for young individuals navigating the complexities of living with HIV.²⁰

The transitional phase from adolescence to young adulthood is often characterized by exploratory behaviors, risk-taking tendencies, and the intricate navigation of multi-faceted decisions pertaining to romantic relationships, sexual behavior, substance usage, and identity formation.²⁰ For individuals in the youth and emerging adult category (aged 18–29 years), these complexities are further compounded by the need to negotiate life while managing a chronic and stigmatized medical condition. For example, HIV stigma scores are most pronounced among individuals aged 18–24 years, closely followed by those in the 25- to 34-year age bracket. Furthermore, the maturation of cognitive processes, marked by concrete thinking and partially developed abstract reasoning, may impede medication adherence among asymptomatic youth, particularly when medications introduce adverse side effects and accentuate their divergence from peer norms.³⁰

This study represents a full-scale, randomized, controlled trial conducted to ascertain the efficacy of the positive STEPS intervention, encompassing various methodological components that enhance both the rigor and potential replicability of the findings. Methodological considerations incorporated within the study encompassed the utilization of randomization, interviewer-assisted and audio computer-assisted self-interviews for the evaluation of adherence, substance use, and sexual behavior, as well as fidelity monitoring and counselor supervision.

Furthermore, the study employed a biological primary outcome measure to assess viral suppression. Nevertheless, it is crucial to acknowledge that this study was not without limitations. In the methodology employed, audio computer-assisted self-interviews were used for capturing responses to sensitive questions and reducing potential biases in self-reported data, particularly in instances where individuals may be inclined to present a more favorable profile. Nonetheless, a noteworthy consideration pertains to the potential influence of demand characteristics, whereby participants in the positive STEPS may have been prompted to artificially inflate their self-reported ART adherence levels. Notwithstanding this concern, the measurement tool employed in this study evaluated adherence through specific questions that are widely recognized in ART adherence trials. Moreover, the implementation of interviewer-assisted and audio computer-assisted self-interviews within a private setting may have effectively mitigated the likelihood of differential reporting. Moreover, the individuals who opted to participate in the study may not accurately reflect the broader population of young people living with HIV. To improve the generalizability of the findings, inclusion of more than 2 study sites would have been advantageous, considering the potential variations in characteristics across diverse geographical locations within the United States. Additionally, the overall retention rate across all follow-up assessment visits was 81%, posing a potential source of bias if the characteristics of retained participants differed significantly from those lost to follow-up. It is noteworthy that the level of attrition observed was within the anticipated range, as accounted for in the study’s statistical power calculations, and is a common occurrence in interventional studies with a 12-month follow-up period. Given that attrition rates did not vary by study condition, it is probable that any resulting bias in the outcomes would lean toward a null effect, indicating a likelihood of underestimating the effect sizes. Consequently, the attrition experienced may have led to more conservative estimations of the true intervention effects.

Moving forward, it is imperative to further investigate effectiveness and implementation science approaches to establish the real-world applicability of positive STEPS across various settings (eg, HIV clinics vs. community-based HIV service organizations vs. hospitals); diverse geographical locations (eg, rural vs. urban); assorted delivery modes (eg, remote vs. in-person participation and delivery by a trained counselor vs. using technology/an app or online platform); and varied levels of counselor training (bachelor’s degree vs. master’s degree), as well as maintenance of gains extending >12 months. These findings of the positive STEPS intervention in significantly enhancing ART adherence and achieving viral suppression—when compared with a SOC control—among youth living with HIV provides a very promising and efficacious approach for addressing this problem in the field of adolescent HIV prevention and treatment research. This study represents highly favorable findings, being the first of its kind to demonstrate such impressive outcomes as far as viral suppression among youth living with HIV in the United States. The results not only showcase the efficacy of positive STEPS but also highlight its ability to sustain these gains over a 12-month period.